Levered vs Unlevered Beta: How to Re-Lever Beta for a DCF, and Why the Raw Number off Bloomberg Is the Wrong One to Use
Michael King, PE Investment Manager · 9 min read ·
- The beta you observe on Bloomberg or FactSet is a levered (equity) beta — it measures a stock’s sensitivity to the market, but that sensitivity is inflated by the company’s own debt. Two identical businesses with different leverage will show different betas even though their underlying business risk is the same
- Because leverage is baked in, you cannot simply average the raw betas of a peer group that carries different debt loads. You must first unlever each peer’s beta — strip the leverage out to isolate pure business risk — then average those, then re-lever at your own company’s capital structure
- The mechanic is the Hamada equation: unlevered β = levered β ÷ [1 + (1 − t) × D/E], and re-levering runs the same formula in reverse. For a private company with no observable beta of its own, this is the only way to get a defensible one
- The step is not cosmetic: getting it wrong — averaging raw betas, or re-levering at the wrong capital structure — typically moves the cost of equity by 50–100bps, which flows straight into WACC and can shift a DCF value by a tenth or more
Beta Measures Risk, But It Carries Two Kinds at Once
Beta is the number in the cost of equity that scales the equity risk premium to a specific company. A beta of 1.0 means the stock moves with the market; 1.5 means it swings 50% harder in both directions; 0.7 means it is more defensive than the index. That much most candidates can recite. The part that separates a clean answer from a weak one is knowing that an observed beta is measuring two distinct risks fused together: the business risk of what the company actually does, and the financial risk added by how much debt it carries.
Debt amplifies equity returns in both directions — a leveraged company’s shareholders sit behind a fixed interest bill, so the same swing in operating profit produces a larger swing in what is left for equity. That amplification shows up as a higher beta. The consequence is immediate: the beta printed on a Bloomberg screen — the levered, or equity, beta — is specific to that company’s balance sheet, and it is the wrong number to lift straight into a valuation of a company with a different balance sheet.
The Contamination Problem: A Peer’s Debt Is Sitting Inside Its Beta
The reason this matters in practice is that betas come from comparable companies. You rarely value a business against its own long price history — for a private company or an LBO target there isn’t one — so you build a peer group of listed comparables and read their betas. But those peers were assembled to match business risk: same sector, same end markets, same cyclicality. They were not assembled to match capital structure, and they almost never do.
The fix is to convert every peer’s levered beta into the beta it would have with no debt at all — the unlevered, or asset, beta — which strips out the financial risk and leaves only the business risk the peer group was chosen to represent. Once every peer is expressed on that common, debt-free footing, the numbers are finally comparable and can be averaged honestly.
Unlevering: Stripping the Leverage Out
The standard tool is the Hamada equation, and in interview conditions you are expected to know it cold:
Run the three peers through it at a 25% tax rate — the current UK corporation tax rate, so the natural default for a London-based process:
| Peer | Levered β | D/E | Unlevered β |
|---|---|---|---|
| Peer A | 1.30 | 0.25 | 1.30 ÷ [1 + 0.75 × 0.25] = 1.09 |
| Peer B | 1.50 | 0.60 | 1.50 ÷ [1 + 0.75 × 0.60] = 1.03 |
| Peer C | 1.10 | 0.10 | 1.10 ÷ [1 + 0.75 × 0.10] = 1.02 |
Notice what the unlevering reveals. On a raw basis the peers looked spread out — 1.10 to 1.50, a spread of 0.40 in beta terms. Stripped of leverage they collapse to 1.02–1.09, a spread of 0.07. That tightening is the whole point: almost all the apparent difference in their betas was financial risk, not business risk. The three companies are, as businesses, far more alike than their headline betas suggested. Average the unlevered figures and you get a clean business-risk beta of roughly 1.05.
Re-Levering: Putting Your Own Capital Structure Back In
The 1.05 asset beta describes the business with no debt. Your target will have debt — possibly a great deal of it, if it is an LBO — so the final step re-introduces leverage, this time at your company’s structure, by running the equation in reverse:
Say the target runs at a D/E of 0.50, more leveraged than any single peer. Re-lever the 1.05 asset beta: 1.05 × [1 + 0.75 × 0.50] = 1.05 × 1.375 = 1.44. That 1.44 is the beta that belongs in the CAPM — it carries the peer group’s business risk re-expressed at the target’s financial risk. Compare it to the 1.30 you would have used by lazily averaging the raw betas, and the gap is 0.14 of beta.
Why This Feeds Straight Into WACC and the DCF
The re-levered beta is not the destination; it is an input two steps back. It goes into CAPM — cost of equity = risk-free rate + β × equity risk premium — to produce the cost of equity, and the cost of equity is the larger, more sensitive half of WACC. WACC is the rate that discounts the unlevered free cash flows in the DCF. An error in beta therefore does not stay contained: it propagates through the cost of equity, into WACC, into every discounted cash flow and the terminal value, and the terminal value is usually 60–80% of the total. A beta that is too high by 0.14 quietly shaves a tenth off the valuation, and does it invisibly, because every downstream number still looks reasonable.
That propagation is exactly why interviewers linger on this step. It is a compact test of whether a candidate understands that a DCF is a chain, and that a defect near the front of the chain is the most dangerous kind. Which is where the traps come in.
The Four Traps Interviewers Set
The mechanic is memorisable in an afternoon. What separates candidates is handling the judgement calls the formula hides — and interviewers know precisely where those are.
The Verdict: Precise Machinery Bolted Onto Imprecise Inputs
The unlever-average-relever routine is worth doing and worth doing correctly — skipping it, or averaging raw betas, is a genuine error that misprices risk by a knowable margin. But it is worth being honest about what the exercise is. It is exact arithmetic wrapped around three estimates that are anything but: which peers belong in the set, what the target’s real long-run leverage is, and what the equity risk premium even is — a figure the market cannot agree on to within a full percentage point. The formula hands back a re-levered beta to two decimals and lends the whole valuation an air of precision the inputs do not support.
The candidate who executes the mechanic flawlessly but cannot defend the peer set is polishing the third decimal while the first is a guess. The one who says “here is the beta, here is the structure I re-levered to and why, and here is how much the answer swings if I drop this comp or move the risk premium 50bps” is doing the actual job. Get the machinery right, then remember it is only as good as the three judgements feeding it — that combination, not the algebra alone, is what a good interviewer is grading.
Take Your Preparation Further
Beta is one link in the discount-rate chain, so read it alongside the rest. See where the re-levered beta lands next in the WACC calculation guide; follow the cash flows it discounts in unlevered free cash flow; watch the whole thing come together — and see why the terminal value carries most of the weight — in the DCF terminal value guide; and for the peer-selection judgement that decides the asset beta before any formula runs, work through trading comps and precedent transactions and how to think about valuation.
For every valuation method on one page — DCF, comps, precedent transactions and the multiples that anchor them — download the free Valuation Methods Cheat Sheet, and to build the full model with the beta step wired in, the Professional DCF Model Template.
Ready for personalised feedback? Book a 1-on-1 mentoring session with an experienced IB/PE professional.
Frequently asked questions
What is the difference between levered and unlevered beta?
Levered beta (also called equity beta) is the beta you observe on a data provider like Bloomberg — it measures a stock’s sensitivity to the market, but that sensitivity is inflated by the company’s debt, because leverage amplifies equity returns. Unlevered beta (also called asset beta) strips the debt out to isolate the pure business risk of what the company does, independent of how it is financed. Two companies in the same business with different amounts of debt will show different levered betas but similar unlevered betas.
Why do you unlever and re-lever beta in a DCF?
Because betas come from comparable companies, and those peers carry different amounts of debt than each other and than your target. Averaging their raw levered betas would import their financial risk, not just the business risk you actually want. So you unlever each peer’s beta to put them all on a common debt-free footing, average those asset betas, then re-lever the average at your own target’s capital structure. That produces a beta that reflects the peer group’s business risk expressed at the target’s specific financial risk — the correct input for the cost of equity.
What is the formula to unlever and re-lever beta?
The standard tool is the Hamada equation. To unlever: unlevered beta = levered beta ÷ [1 + (1 − t) × (D/E)], where t is the marginal tax rate and D/E is the peer’s debt-to-equity ratio at market value. To re-lever, run it in reverse: re-levered beta = unlevered beta × [1 + (1 − t) × (D/E target)], using the target company’s own debt-to-equity ratio. The (1 − t) term reflects the tax shield on debt. Note the standard formula assumes the debt itself has a beta of zero, which is reasonable for investment-grade balance sheets but less so for heavily leveraged ones.
Do you re-lever beta at the current or target capital structure?
For a stable public company, use the current capital structure. For an LBO or any company whose leverage is expected to change materially, re-lever at the intended long-run (steady-state) capital structure rather than the spike of debt at closing — that day-one leverage amortises down over the hold, so using it would overstate beta for the whole forecast period. Naming which structure you re-levered to, and why, is one of the things a strong interviewer is specifically listening for.
How much does getting beta wrong actually change a valuation?
Materially. In a typical peer set, the gap between a correctly re-levered beta and a naive raw-average beta is often around 0.10–0.15 of beta, which at a 5% equity risk premium is roughly 50–100bps of cost of equity. That difference flows straight into WACC and discounts every cash flow and the terminal value — and because the terminal value is usually 60–80% of a DCF, a 70bps move in the discount rate can shift the equity value by well into double-digit percentages. It is not a rounding step.